very comparable values for end-diastolic and end-systolic volume as those calculated with magnetic resonance images in the same mouse (compare Figure 9 withFigure 17). 255 Knowlton, Kirk U. A DesmoplakinKO RV EDV= 81 ul Segmentation of 3D geometric model Data RV endocanfium of RV endocanfim B Wild RV EDV = 40 ul Segmentation of Data 3D geometric model RV endocanfium of RVendocardium Figure 17: Contrast echocardiographic images for right ventricular geometry and 3Dreconstruction The cross-sectional images utilized for our 3D reconstructions will likely be improved by the use of the Vevo 770 , as noted above. C.3.3. Magnetic Resonance Imaging The new fMRI facility at UCSD is currently being using for mouse heart imaging. Anatomic images are routinely acquired, with other capabilities such as tissue tagging for regional strain measurements, and DTMRI for myocyte structural analysis. These murine imaging protocols will be performed on the 7T horizontal-bore MR scanner (Varian with a new General Electric console), equipped with a 12 cm bore gradient system capable of 22 G/cm gradient strength and a 300 \LS rise time at complete switching. High field imaging is necessary to resolve the sub millimeter scaled anatomy of the mouse heart. A home built dual quadrature transverse electromagnetic mode coil will be used for transmission and reception of the signal at an inner diameter of 1.9 cm. The rapid falloff of magnetic and radio frequency field requires close proximity of the hardware to the imaging specimen. A major advantage of the MR system is its intentional design for small animal imaging. MRI experiments will be conducted using an ECG triggered Fast Low Angle Shot (FLASH) Gradient Echo (GE) pulse sequence. Such a sequence is preferred for high resolution, time sensitive imaging (26). Although susceptible to a more rapid T2* decay due to the influence of magnetic field inhomogeneity, the GE pulse sequence is absent of the 180[unreadable] refocusing pulse that is present in a spin echo (SE) sequence. This concession substantially increases speed of imaging and is a valid compromise. Partial k-space acquisition is employed in the readout direction, whereby a percentage of k-space is sampled per phase encoding step. The reduction of this slice selecting RF pulse itself contributes to a shorter TE and comes at the expense of slice profile. To reduce RF interference between adjacent slices, slice order may be interleaved. To maximize the contrast to noise ratio, the flip angle has been optimized to provide the greatest delineation between myocardium and blood, while not amplifying any oblique flow related artifacts. High resolution slices are readily obtainable in the mouse heart with the following prescription protocol: TR = 5ms;TE = 1.2 ms;66% partial echo;field of view = 25 mm;data matrix = 128;slice thickness = 1 mm;spectral width = 32 kHz;flip angle = 20[unreadable];1 ms RF pulse;and time series averages = 4. Mice will be imaged in vivo under free breathing inhaled isoflurane anesthetics. Anesthetic induction will be with 5 Vol% isoflurane in 100% C>2. After the proper plane of anesthesia is reached, the mice will be transferred into a custom built restraint 256 Knowlton, Kirk U. unit that centers within the RF coil while being maintained with 1Vol-%isoflurane at 1.0 -1.5 l/min. The EGG (for triggering) and core body temperature are monitored;a tail cuff will record systolic blood pressure. Bore temperature is regulated using heated airflow to maintain mice at 36 - 38[unreadable]C, and heart rate is typically 480-520 beats per minute. B Slice 1 Sli[unreadable]t2 Slice 3 Sice 4 Slice5 Slice 1 SlioeS Slice 3 Slice* Slice5 Prolate Spheroidal Mesh 3D Geometric Model Figure 18: Outline of methods for reconstruction of anatomically accurate 3D models of the mouse heart. MR scout images and axial slices are shown in panels A and B. Multiple long-axis views are taken, panel C, and surfaces outlined, panel D, for input into a computational model for fitting, panel E, and 3Drendering, panel F. 257 Knowlton, Kirk U. Frames that capture a particular phase of the cardiac cycle may be readily obtained by employing gating schemes that initiate data acquisition relative to the ECG trace. Retrospective gating methods may be used to continuously acquire data at a high rate, and in a post-processing step, correlate that with the appropriate phase during the cardiac cycle creating a cine array of images that cover an entire R-R interval. Temporal resolution in such schemes is limited by the shortest applicable repetition time, which can be though of as the quantized time necessary to collect a single unit of data. Fast gradient echo pulse sequences have been tailored in a retrospectively gated scheme allowing the cardiac cycle to be parsed into 11-12 phases (27). Figure 18 outlines the steps and results for MR imaging and model reconstruction. Figure 17A is a localizer image, which is a coronal slice through the long axis of the LV. From the coronal slice, an axial slice is taken through the middle of the ventricle (red-dotted line in Figure 17A). Figure 17B shows the axial slice. From the axial slice, five longitudinal slices (red-dotted lines) through the long axis of the LV are prescribed. The slices are separated by 36[unreadable]. All five images are shown in Figure 17C. Using an edge detection algorithm, the LV endocardium and epicardium are defined (Figure 17D). The LV endocardium and epicardium data points are input into our finite element analysis package (Continuity) for anatomical modeling (described in detail below). A prolate spheroidal mesh is fitted to the data to create a 3D geometric model of the LV (Fig. 17E and 17F). From this model, mass, volumes, ejection fraction, wall thickness, and curvature can be calculated. In addition to function obtained from geometric alteration, regional function can be estimated by tagging a MR image. Saturation pulses are applied that tag portions of the myocardium while leaving other sections unaffected. These magnetizing pulses can be configured to render tag lines in both in-plane directions, generating a grid of intersecting points throughout an image that serve as material markers. Since these tag lines are produced from the myocardium itself, they will deform as the heart contracts and relaxes during a cardiac cycle. Strain fields can thus be estimated by tracking the movement of these material points from one cardiac phase to the next. Zhou et al. applied a SPAMM preparation to their cine anatomical sequence to assess global cardiac function (26), and these same techniques will be applied in this proposal. In this Program Project a number of models of right ventricular dysfunction are to be studied. The right ventricle can also be reconstructed and modeled from MR images in addition to the contrast echocardiographic and angiographic methods described above and below. For example, short-axis images can be taken from the base to the apex (Figure 19A). From these images, the right and left ventricle are easily segmented (Figure 19B). The data from these images can be used to construct a biventricular model in our software analysis program (Figure 18C). A B C Short-am MR image Segmentationof 3Dgeometric model RV and LV of ventricles Figure 19: MR images (short axis) for right and left ventricular anatomy and 3D reconstruction. 258 Knowlton, Kirk U. C.3.4. X-ray Video Equipment and Methodology For purposes of angiography mice (andother rodents), we are utilizing the XiScan 1000C-arm x-ray system (XiTec, Inc.,Windsor Locks, CT). This system is presently approved for human use and delivers x-rays in the 40 kV to 70 kV range. The portable C-arm unit incorporates a small x-ray tube and a CCD video camera that records the x-ray image from the input phosphor (3" field of view). The system provides continuous x-ray exposure using an automatic exposure rate control that algorithmically sets values of kilovoltage and microamperage to optimize image quality. Values for kilovoltage and microamperage can also be set manually and held constant throughout an imaging sequence. An automatic brightness control sets brightness and contrast on both display monitors to best utilize the grey scale qualities of monitor screens. Edge enhancement and continuous frame averaging can be used to improve the visibility of fine points in the image, reduce statistical noise, and highlight interfaces between anatomic segments. Images are stored on a super- VMS video tape at 60 fields/s or 30 frames/s. With this system, digitized images also can be stored on a floppy disk for later viewing and processing, and digital subtraction methods can be applied. Presently, for analysis of images obtained during intravenous contrast medium injection, the videofields are Anterior-Posterior Lateral End-Dlastollc Volume: 84.2 Ml;End-Systolic Volume: 57.7 pi Ejection Fraction: 0.31 Figure 20: Anterior-posterior (AP) and lateral (Lat) projections of the right ventricle, visualized by contrast microvideoangiography. The area outlined in yellow represents the assigned region of interest (ROI) at end-systole (ES);that in green represents the ROI at end-diastole (ED). The calculated volumes at ES and ED are computed using Simpson's rule ( ). RVglobal ejection fraction(RV-EF) is [EDV-ESV] /EDV. 259 Knowlton, Kirk U. digitized from S-VHS tape and the image sequence is written onto an ultra SCSI disk drive capable of an internal formatted transfer rate of 59-188 megabytes/s and with a total storage capacity of 300 gigabytes. These images are then edited and digitally processed using a Silicon Graphics R10000 system, running the Motif 6.5 operating system. Before final storage for subsequent processing, each videofield is converted to a videoframe, using pixel-by-pixel interpolation of contiguous raster lines, and thereby providing a temporal resolution of 60 frames/s. The software for this image processing was developed internally and incorporates some routines provided by Silicon Graphics, Inc. Following image digitization, early-frame digital subtraction, and application of Simpson's rule to assigned regions of interest, the end-diastolic and end-systolic volumes can be calculated, Figure 20. The microvideoangiographic approach has also been used to demonstrate pronounced aortic tortuosity in DANCE knockout mice, Figure 21. Figure 21: Tortuous aorta in DANCE-deficient mice. Wild-type mouse is shown on left; knockout mouse is shown on right. C.3.5. Intravital Microscopy Methods in Embryos Intravital studies of embryonic mice by color and fluorescence microscopy of the cardiac chambers has been previously used in this laboratory and will be available if required. Maternal anesthesia is provided with ketamine/xylazine;a midline abdominal incision is performed with subsequent delivery of the embryos through a uterine incision while preserving the placental circulation. At ED10.5-12.5 the heart is visualized by color videomicroscopy in the intact embryo (9). In older embryos, using microsurgical techniques the chest of the embryo is opened, and contrast with fluorescein-labeled albumin is necessary to visualize the LV chamber using left atrial injection of fluorescein-conjugated albumin (9,10). Continuous transillumination of the heart from below is accomplished with a DC tungsten-halogen light source and images of the beating heart are obtained with an intravital microscope. Cardiac images are obtained by a color coupled charge device television camera. C.3.6. Microsurgical Methods Adult mice (transgenic, wild type controls, and sham-operated mice) (8 weeks of age or older) are anesthetized with i.p. ketamine (100mg/kg) and xylazine (5 mg/kg), and under a dissecting microscope animals are intubated and placed on a rodent ventilator, as described (11,12). 260 Knowlton, Kirk U. Transverse aortic constriction (TAC): The chest is opened with a small incision in the 2nd intercostal space and the aortic arch is identified. A 7-0 suture ligature is placed around the transverse aorta and tied around a 24 to 27G needle (size depending on animal weight and on degree of baseline LV dysfunction). The needle is then rapidly removed to produce a pressure load on the LV. The chest is then closed and the pneumothorax evacuated. After 1 to 3 weeks (or longer or shorter periods), and prior to terminating the experiment, pressures are measured simultaneously in the right and left carotid arteries by cannulation with flame-stretched polyethylene tubing using standard strain gage transducers, and the pressure gradient across the stenosis due to the band is measured (11,12). Pulmonary artery banding: The pulmonary artery is identified through a small incision in the 2nd intercostals space, and a suture ligature is placed around the vessel;the suture is then tied against either a 25G (moderate stenosis) or 26G needle (severe stenosis) which is then rapidly removed (14). The chest is then closed and the pneumothoraxevacuated. At the end of the banding period, mice are anesthetized as above and echocardiographic hemodynamics or other studies performed. Hearts are then excised and the atria, right ventricle free wall, septum and left ventricle are carefully dissected under the dissecting microscope and weighed separately. Tissues can then be frozen separately in liquid N2 for later analysis, or perfusion fixed for other studies. Coronary artery ligation: Ligation of the left anterior descending coronary artery in rats and mice produces an anterior-lateral transmural infarct of varying size. Echocardiography is used at 1 week to identify mice with infarcts that are sufficiently large to cause LV dilation and hypertrophy (30-40% of LV chamber perimeter), which usually occurs by 3-4 weeks;larger infarcts (>40%, of LV perimeter) typically produce LV dysfunction at 1-2 months. The LV ejection fraction is measured by echocardiography using multiple chords perpendicular to the chamber long axis to calculate LV volume by Simpson's rule. C.3.7. C.3.8. Methods for In Vivo Gene Transfer Intraaortic Injection During Hypothermic Cardiac Arrest. Animals are anesthetized with ketamine/xylazine and ventilated. After disinfecting the anterior chest, a small (<1 cm) left thoracotomy is performed through the 2nd intercostal space. Sutures (5/0 silk) are looped around the ascending aorta and pulmonary artery and threaded through flared plastic occluded tubes. A flame-stretched 5 cm length of PE-50 tubing is used to measure aortic pressure and advanced from the right carotid artery into the aortic root, just above the aortic valve. The animal is then immersed in ice water while monitoring rectal thermometer, with cooling to 25-26[unreadable]C accomplished quickly in mice (approximately 10 minutes). The pulmonary artery is first occluded and ascending aorta is then occluded by the snares placed above;the tip of the tubing in the aorta is verified to be proximal to the aortic snare by brief aortic constriction. Solutions including modified cardioplegic solution and histamine, or substance P, followed by viral vector injection are delivered within 2 to 3 minutes. The clamps on the great vessels are then released, dobutamine administered and rewarming commenced. After removal of air, the chest is closed, the animal extubated, and allowed to recover (16). Following transthoracic echocardiography at 4-5 days (5 weeks or longer with AAV) after the operation, animals are euthanized by an overdose of sodium pentobarbital (100 mg/kg) and the heart, kidney, lungs and liver are quickly removed, embedded in OCT compound (Miles, IN), frozen in isopentane precooled with liquid nitrogen and stored at -80[unreadable]C. To estimate the efficiency of nuclear targeted p-galactosidase gene transfer, tissues are incubated with Bluo-Gal (Life technology) buffer (blue stained cell nuclei considered positive);point counting is then done of stained and unstained nuclei. Intravenous Injection Without Hypothermic Arrest. Mice are anesthetized using a mixture of 1.5% isoflurane and oxygen (1 to 2 L). Following isolation of the jugular vein, a 29-gauge sterile needle and syringe are then used to deliver recombinant AAV9 viral vectors via the vein in a volume of 150 uL (17). At 6-8 weeks following 261 Knowlton, Kirk U. the viral injection, and following transthoracic echocardiography, the animals are euthanized, and the tissues examined, as stated above, for efficiency of nuclear targeted p-galactosidase gene transfer. C.4 CORE PERSONNEL Kirk L. Peterson, M.D. (Core-Leader) has an extensive background in clinical and experimental cardiac catheterization methods, hemodynamics, and cardiac imaging. He now has accumulated extensive experience with phenotypic assessments in large and small animals, as well as gene transfer methods in vivo. He will bring skills in developing and applying new computer programs and technology. He also is experienced in data quality control, data analysis, development of animal models, and experimental design. Jeff Omens, Ph.D. (Core Co-Leader) is an expert in regional function analysis in the small animal heart. He has over 20 years experience with large and small animal models of cardiac disease. He has also been responsible for implementation of the UCSD fMRI facilities for cardiac imaging. He developed and employed several imaging techniques for isolated and intact murine cardiac function, as well as computational models of cardiac mechanics. He will be directly involved with experimental design, data acquisition and analysis, and development of new imaging and modeling techniques. Masa Hoshishima, M.D. (Collaborating Investigator) is an experienced molecular biologist and also has a strong background in developing and applying viral vectors in vivo. Yusu Gu, M.S. (Master of Medicine, Collaborating Investigator) has had training in physiology and molecular biology;she is highly skilled in performing microsurgery in mice. She supervises and performs physiologic studies, carries out aortic banding procedures with a high success rate, uses high fidelity micromanometer catheters and conductance catheters in mice, and is capable in a variety of tissue analyses (immunostaining, dye injections, TUNEL staining, and Northern blot analysis), and has been integral to the development of methods for in vivo gene transfer. Nancy Da/ton is a highly experienced echocardiographer who has participated actively in developing our methods for echocardiography in mice. She performs most of our echocardiographic studies for transgenic colony screening, as well as 2D and M-mode studies in normal, transgenic, and cardiac overload models in the mouse. Far/of Abdel-Wahhab is manager of the Core laboratory. He is experienced in setting up and participating in a variety of experiments. In addition to maintaining supplies and equipment, sterile instruments and pharmaceutical supplies, he is experienced in animal care. He also trains other support personnel and new investigators in laboratory procedures. D. FACILITIES AVAILABLE All equipment and facilities requested for this Program Project Renewal Application are currently available to Core B and are described at the end of this Core Unit. E. BUDGET/BUDGET JUSTICATION See Budget pages preceding Core B. F. RELATION OF CORE UNIT TO RESEARCH PROJECTS All of the research projects will use Core B. Unit 1 (Knowlton) will utilize this Core for physiologic characterization of adult coxsackie-adenovirus receptor (CAR) mutant mice, utilizing echocardiography, hemodynamic, and angiographic procedures. These same mice will also undergo a left ventricular pressure overload (transaortic constriction) in order to establish whether CAR is important in the development of typical pressure overload ventricular hypertrophy, or, whether its absence leads to a dilated cardiomyopathy. Further, Unit 1 will need the resources of this Core for realization of its aim to establish the specific role of an intracellular binding partner of CAR, known as ZO-1, on left ventricular systolic and diastolic function. Unit 2 (Chen) will access the personnel and resources of Core B in order to analyze the in vivo physiological and morphological characteristics of the mouse mutant intended to recapitulate Naxos disease, a recessive form of 262